Project Summary/Abstract Veterans and other Americans who survive stroke often face disabling motor impairments that impede performance of activities of daily living and limit free-living activity. Prominent among these are diminished locomotor function and impaired balance that not only foster a sedentary lifestyle and physical deconditioning, but also increase the risk injuries due to falls. Recent research from our group at the Baltimore VAMC and others around the world has demonstrated how motor learning based interventions can modify brain activity and improve motor functions in persons with stroke. Now there is a major research opportunity to advance the effectiveness of these interventions by applying new robotics technologies to improve neuromotor control of essential functions such as gait and balance. One critical area for performance of walking and standing balance is the control of the ankles, as they are a major conduit of mechanical power in gait and also modulate torques affecting the motion of the whole body center of mass when balancing. Thus the current proposal is designed to investigate two approaches for using an impedance controlled ankle robot to improve gait and balance function among stroke survivors with chronic lower extremity hemiparesis. One approach uses the ankle robot in a seated visuomotor training program that focuses on improving paretic ankle motor control that may transfer to gait and balance functions. The other approach follows the dominant rehabilitation paradigm of task-specific training by integrating use of the ankle robot during treadmill exercise training to assess effects on the same outcomes. The effectiveness of both robotics approaches will be compared to that of a treadmill exercise program without robotics. The study tests the hypothesis that, in persons with chronic lower extremity hemiparesis, 6 weeks of seated ankle robot training will improve paretic ankle motor control with major improvements in standing balance and moderate improvements in gait, whereas the same amount of training on the treadmill with the ankle robot will improve gait function more than balance. Both robot-trained groups will outperform the treadmill only group on balance, while the treadmill + robot group will make the greatest gains in gait and the seated robot group will make some improvement in gait but will show greater gains in ankle motor control and balance. Aims: In a 6-week intervention (18 sessions) with persons with chronic lower extremity hemiparesis 1) Compare effects of seated visuomotor ankle robot training vs. treadmill + robot training on paretic ankle impairments and motor control; 2) Compare effects of seated-robot vs. treadmill + robot training on functional mobility and balance outcomes; and 3) Compare the effectiveness of both robotics approaches to a standard treadmill exercise protocol of the same duration. This proposal will establish the initial comparative efficacy of two motor learning based approaches using a modular impedance controlled ankle robot and contrast motor control and functional gait and balance outcomes among them. As a pilot study we will establish initial deficit profiles for users that respond to each intervention across the 6-week period. This will begin to fill the knowledge gap on how to apply robotics technologies in the elderly stroke population. Such findings have the potential to change the paradigm of stroke care based upon advances in our understandings of activity-dependent brain plasticity and the critical need for improving mobility in the rehabilitation of Veterans and others with disabilities. Our results will provide the requisite information to design a larger Merit Award trial that refines or combines these interventions to reduce risk factors for disabilities and limited social participation in the growing stroke population.